This paper considers individual differences in the Emotion Recognition Ability (ERA) of 1368 participants in different modalities. The sample consisted of 557 first language (L1) and 881 foreign language (LX) users of English from all over the world. This study investigates four independent variables, namely modality of communication, language status (L1 versus LX), proficiency, and cultural background. The dependent variable is a score reflecting ERA. Participants were asked to identify an emotion (happiness, sadness, anger, fear, surprise and disgust) portrayed by a native English-speaking actress in six short recordings - either audiovisual or audio-only - embedded in an online questionnaire. English proficiency was measured through a lexical recognition test. Statistical analyses revealed that participants were better able to recognise emotions when visual cues are available. Overall, there was no difference between L1 and LX users' ERA. However, L1 users outperformed LX users when visual cues were not available, which suggest that LX users are able to reach Ll-like ERA when they can rely on a sufficient amount of cues. Participants with higher proficiency scores had significantly higher ERA scores, particularly in the audio-only condition. Asian LX users were found to score significantly lower than other LX users.
